Psychotic disorders in adolescents present acute and long-term challenges, including those of differential diagnosis, prognosis and treatment choices (Kafantaris et al. 1996, 1998). When lithium became widely available in the U.S., much attention was paid to differential diagnosis of mood disorders with psychotic features from schizophrenia because, for some patients, lithium was more effective than first-generation antipsychotics (Janicak et al.1992). Second generation antipsychotics (SGAs) have been shown to be efficacious as monotherapy for the treatment of acute mania (Perils et al. 2006) and to increase response rates when added to a traditional mood stabilizer, including in adolescents with or without psychotic features (Del Bello et al. 2002). In a small study, adolescents who had prominent psychotic features at study entry appeared to benefit from continuing adjunctive treatment with an antipsychotic agent despite ongoing lithium treatment (Kafantaris et al. 2001a, 2001b). As many of the treatments approved for bipolar disorder now converge with those used for schizophrenia, pinpointing the exact diagnosis, particularly in adolescents, may seem less crucial and call into question whether treatment with traditional mood stabilizers for bipolar disorder with psychotic features is even necessary. Moreover, although the addition of valproate to a first-generation antipsychotic resulted in a higher acute response rate in one adult study (Muller-Oerlinghausen et all 2000), it is not known whether it is beneficial to add a traditional mood stabilizer to a SGA for acute and continuation the treatment of psychotic mania. This is an important question, particulariy for adolescents, as the treatment of an early psychotic episode has significant impact on course of illness and developmental outcome. The proposed study is a placebo-controlled, parallel group, randomized clinical trial comparing two treatment strategies in adolescents with mania and prominent psychotic features. One group will receive a SGA and placebo and the other will receive a SGA and lithium. During the acute treatment phase, the possible benefits of adding lithium to a SGA include a higher proportion of responders, shorter time to response, and potentially a lower dose of SGA. For continuation treatment, possible benefits include prophylaxis against episodes of depression. Results from this study will inform the design of a more definitive trial. Specific aims are: 1. To assess the efficacy and safety of adjunctive lithium in the acute treatment of mania with psychotic features in adolescents. 2. To assess the efficacy and safety of adjunctive lithium in the prophylaxis against recurrence of mood episodes with psychotic features in adolescents over a six month follow-up period.